We are conducting a collaborative case-control study to identify and evaluate risk factors for multiple myeloma and chronic lymphocytic leukemia. Cases of these diseases are being identified through population-based cancer registries in Washington State, Utah, Detroit and Atlanta over a three-year period. Control subjects are being identified from the general populations from which the cancer cases arose. Probability sampling methods, such as area sampling and random digit dialing, are being used for selection of controls. Cases and controls are being interviewed with particular attention to variables related to chronic antigenic stimulation, e.g. past illnesses, allergies, drugs, etc. We are also obtaining detailed residence and occupational information in an attempt to generate new etiologic hypotheses.